T. Blocker et al., The rapidly evolving hypergiant IRC+10 420: High-resolution bispectrum speckle-interferometry and dust-shell modelling, ASTRON ASTR, 348(3), 1999, pp. 805-814
The hypergiant IRC +10 420 is a unique object for the study of stellar evol
ution since it is the only object that is believed to be witnessed in its r
apid transition from the red supergiant stage to the Wolf-Rayet phase. Its
effective temperature has increased by 1000-2000 K within only 20 yr. We pr
esent the first speckle observations of IRC +10 420 with 73 mas resolution.
A diffraction-limited 2.11 mu m image was reconstructed from 6 m telescope
speckle data using the bispectrum speckle-interferometry method. The visib
ility function shows that the dust shell contributes similar to 40% to the
total flux and the unresolved central object similar to 60%.
Radiative transfer calculations have been performed to model both the spect
ral energy distribution and visibility function. The grain sizes, a, were f
ound to be in accordance with a standard distribution function, n(a) simila
r to a(-3.5), with a ranging between a(min) = 0.005 mu m and a(max) = 0.45
mu m. The observed dust shell properties cannot be fitted by single-shell m
odels but seem to require multiple components. At a certain distance we con
sidered an enhancement over the assumed 1/r(x) density distribution. The be
st model for both SED and visibility was found for a dust shell with a dust
temperature of 1000 K at its inner radius of 69 R*. At a distance of 308 R
* the density was enhanced by a factor of 40 and and its density exponent w
as changed from x = 2 to x = 1.7. The shell's intensity distribution was fo
und to be ring-like. The ring diameter is equal to the inner diameter of th
e hot shell (similar to 69 mas). The diameter of the central star is simila
r to 1 mas. The assumption of a hotter inner shell of 1200 K gives fits of
almost comparable quality but decreases the spatial extension of both shell
s' inner boundaries by similar to 30% (with x = 1.5 in the outer shell). Th
e two-component model can be interpreted in terms of a termination of an en
hanced mass-loss phase roughly 60 to 90 yr (for d = 5 kpc) ago. The bolomet
ric flux, F-bol, is 8.17 . 10(-10) Wm(-2) corresponding to a central-star l
uminosity of L/L. = 25 462 . (d/kpc)(2).